EESM 698HSpring 2011

Handoff Latency and Delay Minimisation in Wireless

Networks

Koushik Kumar NundyMaster of Science (Telecommunications), pursuing

Dept. of Electronics and Computer EngineeringHong Kong University of Science and Technology

under the guidance of

Prof. Vincent K.N. LauProfessor

Dept. of Electronics and Computer EngineeringHong Kong University of Science and Technology

- i -

ACKNOWLEDGEMENTS

This project report has been prepared in partial fulfillment of my

Master of Science in Telecommunications curriculum under the guidance of Prof.

Vincent K.N. Lau. For successful completion of this project and report, I am

extremely thankful to my instructor, without whose kind help this project would not

have come to fruition.

I would also like to thank Prof. M.K.Naskar and Mr. D. Sarrdar of

Jadavpur University, Prof. A. Chandra of NIT Durgapur, and Prof. K.B.Letaief &

Prof. R.Cheng of HKUST for their guidance in the field of telecommunications in

general and handoff management in particular.

For this report I owe my gratitude to the colossal storehouse of

information known as the Internet and also to certain papers & books by several

authors on the topic of wireless handoffs. Also, I am grateful to my universities, NIT

Durgapur and the Hong Kong University of Science and Technology, for the

academic background they have offered me, without which I could neither attempt the

project nor be able to complete it.

___________________________________________________________________

Koushik Kumar Nundy

Hong Kong University of Science and Technology

- ii -

CONTENTS

Topic. Page.

Acknowledgment. i

Table of contents. ii

List of figures and illustrations. iii

Chapter 1 – Abstract. 1

Chapter 2 -Introduction. 2

Chapter 3 -Preliminaries and Definitions 3

3.1 Handoff

3.2 Soft Handoff

3.3 NGWS

3.4 IEEE 802.11

3.5 GPS

Chapter 4 – Background and literature survey 6

4.1 Cross Layer Mobility

4.2 Velocity and Received Signal Strength Estimation

4.3 Location Information and GPS

4.4 Network Graphs

Chapter 5 – Overview of Current Methods 8

5.1 Standard Handoff mechanisms in IEEE 802.11

5.2 Standard Handoff mechanisms in LTE networks

Chapter 6 – Suggested improvements 11

6.1 Modified Neighbour Graph Technique for IEEE 802.11

6.2 Pre-Selective Scanning with GPS in IEEE 802.11

Chapter 7 - Results & Discussion 15

7.1 Modified Neighbour Graph Technique

7.2 Pre-Selective Scanning with GPS

Chapter 8 - Conclusion and Future scope. 17

References. 18

- iii -

LIST OF FIGURES AND ILLUSTRATIONS

Fig 3.1 Soft and Hard Handoff

Fig 3.2 Logo of the Wi-Fi Alliance

Fig 5.1 The Typical IEEE 802.11 Handoff Procedure

Fig 5.2 Handoff flow in LTE without SGW change

Fig 5.3 LTE X2 Based Handover

Fig 6.1 Neighbour Graph Flow diagram

Fig 6.2 Case A with 2 adjoining NBS

Fig 6.3 Case B with 3 adjoining NBS

Fig 6.4 Pre selective Scanning Algorithm

Fig 7.1 Scanning Delay(basic active scan)

Fig 7.2 Scanning Delay(fast active scan)

Fig 7.3 Scanning Delay(basic selective scan)

Fig 7.4 Scanning Delay(selective scanning) using cache

- 1 -

Chapter 1 - Abstract

For wireless networks with mobile stations dependent on a central master

controller (base station/access point), an important aspect of reliable service is in the

maintenance of reliable and good radio link quality between the mobile station and the

fixed controller. For this to suitably happen, we must be able to switch control from one

base station to another at the cell boundary to maintain the quality and existence of the

call. A major bottleneck in ubiquitous, always best service occurs at handoff, with

several instances of false handoff initiations resulting in network congestion and sub-

par utilisation, and of handoff failures, due to inordinate delay in completion of handoff,

resulting in termination of old connection before the new one is established. In this

project we discuss current methods used to mitigate the problems of handoff latency

and false handoffs, and suggest possible improvisation and modifications on curent and

predicted technologies that can offer better handoff performance.

In this report we present techniques for reducing hand-off latency in several

kinds of networks, including Next Generation Wireless Networks (NGWS) and Wi-Fi

Networks(IEEE 802.11).

- 2 -

Chapter 2 – Introduction

Wireless networks have become an inalienable part of human life today. With

nearly every individual owning multiple device that makes use of wireless

communication networks, they have practically moved to ubiquity. Under the

circumstances, it is often found that new devices can either join a framework or

backbone wirelessly, acting as what can loosely be called a client server structure, or

develop an ad-hoc network, with a direct connection between two or more equivalent

nodes.

Handoff is a complicated process, and if improperly executed, may result in

severe data loss, drop in effective channel capacity, and even cause the connection to be

dropped. Thus handoff management is a primary design parameter in any such network

that requires it. This has under it's purview nearly all familiar wireless LAN, MAN and

WAN networks like GSM, CDMA, IEEE 802.11(Wi-Fi), Wi-Max, etc.

There is a need for reduced handoff time in wireless networks , since lower

handoff periods would result in higher data efficiency and fewer handoff failures. This

would cause fewer data packet losses resulting in higher QoS, which is a primary facet

of NGWS networks. Moreover, 4G networks integrate certain microcellular networks,

like IEEE 802.11, which requires quicker handoff because number of handoffs become

exponentially higher and cell size dramatically drops with respect to macrocellular

networks. Also effectively, the time spent in handoff cannot be used for useful data

transfer.

This report presents a method of reducing hand off delay and minimising

handoff failure probability using available technology, without compromising

bandwidth efficiency considerably, by altering and streamlining the functionality and

task division of the contemporary handoff technique. This would result in fewer call

failures and in congestion-free networks.

- 3 -

Chapter 3 - Preliminaries and Definitions

3.1 Handoff

“As a mobile instrument proceeds from one cell to another during the course of

a call, a central controller automatically reroutes the call from the old cell to the new

cell without a noticeable interruption in the signal reception. This process is known as

handoff. ”- Britannica[1].

A better description[2] would be , “In cellular telecommunications, the term handover or

handoff refers to the process of transferring an ongoing call or data session from one

channel connected to the core network to another.”

As we can see the terms handoff and handover can be used interchangeably and

mean the same, the latter being in more common use in European literature.

3.2 Soft Handoff

In this report we will be considering only soft handoffs, where the connection

with the new base station is attempted before the connection with the old base station is

terminated, rather than immediately terminating the connection between a Mobile

Station(MS) and a Base Station (BS). In the course of handoff, a new connection is

established first between the MS and a new BS, while keeping the old connection

between the MS and the old station. Only after the new connection can stably transmit

data, the old connection is released. Thus, Soft Handoff[3] is a “make before break”

mechanism. This mechanism helps ensure the service continuity, which is however at

the cost of more capacity resource consumption during the handoff (as two connections

are established simultaneously).

The difference between soft handoffs and hard handoffs(break before make) can

be more clearly understood from Fig 3.1.

- 4 -

Fig 3.1 Soft and Hard Handoff

3.3 NGWS

Next Generation Wireless System (NGWS)[4][5] consists of different wireless

architectures ranging from cellular wireless networks to satellite networks. It is

basically a network which follows the always best principle, in form of a heterogeneous

network which incorporates networks of various sizes and properties using multiple

wireless communication standards to create a network always accessing the best link

(optimal throughput, delay and reliability) for communication.

3.4 IEEE 802.11

IEEE 802.11 is a set of standards for Wireless Local Area Networks(WLAN) as

specified by the IEEE LAN/MAN Standards Committee. It is colloquially referred to as

Wi-fi.

The 802.11 family consists of various over-the-air modulation techniques which

use the same basic protocol. The latest release of the specifications is 802.11n, which

incorporates enhancements for increased throughput.

The base current version of the standard is IEEE 802.11-2007[6].

- 5 -

Wi-fi is technically a more general de-facto standard based on the IEEE 802.11,

and is specified by the Wi-fi Alliance.

Fig 3.2 Logo of the Wi-Fi Alliance

3.5 GPS

GPS or Global Positioning System is a space based location determination

standard using satellites. Originally developed by the US military in 1973, currently, it

is open for use by everyone, who can receive position(latitude and longitude) data, as

well as time information, using a receiver unit. Similar global navigation satellite

systems include the Russian GLObal NAvigation Satellite System (GLONASS), opened

for civilian use in 2007. Planned alternatives are Compass Navigation system (China),

and Galileo positioning system (EU).

Several mobile devices, notably “smartphones”, have started to incorporate GPS

receiver units, thereby opening up the possibilty of it's use in telecommunication

equipment.

- 6 -

Chapter 4 – Background and literature survey

Considerable work in handoff delay minimisation has been done by Mohanty et

al.[5][7][8], Tekinay[9], Mollini and several others in various types of networks, mostly in

IS-95, WCDMA, WI-Fi, Wi-Max and LTE systems.

According to Shantidev Mohanty[7][8] during any handoff between two base

stations efficient intra and inter system handoff protocols should have limited handoff

latency,low packet loss and limited handoff failure to support seamless roaming.

4.1 Cross Layer Mobility

One of the older examples of using information from multiple layers[8], stated

that, using speed and handoff signaling delay information the performance of existing

handoff management protocols can be enhanced. Mohanty et.al.[7] presented a cross

layer mobility model where the data link layer and the network layer could be used for

speed and handoff signaling delay estimation. It was called Cross Layer Handoff

Management Protocol.

4.2 Velocity and Received Signal Strength Estimation

Narsimhan & Cox shows us a simple technique to estimate the speed of the

mobile station[10], using Doppler Shift and show that location within the cell be

estimated using Received Signal Strength(RSS)[11], to determine the suitable point

where handoff should start. They propose that when combined with an estimation of the

Doppler shift of the signal received, cell boundaries and time required for handoff can

be derived to a reasonable extent[10] [11].

Nundy et.al.[12] discusses how to reduce false handoff initiation in low latency

networks, using speed and signal strength information, which in turn will help reduce

handoff delay and resultant overhead.

- 7 -

4.3 Location Information and GPS

Sarrdar et al.[13] suggested a pre-scanning of adjoining cells to detect available

channels, using GPS based location information. This resulted in faster handoff, and

lower handoff failures without putting any appreciable load on the base station or the

network.

Also, in an early implementation, a patent[14] discusses the use of location

determination for handoff in mobile systems.

Dutta et.al.[15] shows an efficient mechanism for GPS assisted handoffs in real

time communication systems. It provides a new methodology that can provide faster IP

address discovery using GPS coordinates of the mobile station.

Celebi & Arslan[16], apply GPS information in cognitive networks for more

efficient signalling and handoff.

We see other examples of location assisted handoff in the works of Cortes-

Rodriguez et al.[17] and Raith[18], among others.

4.4 Network Graphs

Network mapping and graphs have been used to estimate handoff initiation and

associated signalling for a long time. In recent times, several techniques have been

suggested or implemented that utilise both such graphs and location data, to form a

more efficient handoff scheduling mechanism. This approach is even more helpful in

microcellular and ad-hoc type networks, where the nature of such maps is dynamic and

needs to be varied in real time, thus making node location even more pertinent.

In a recent unpublished work[19], we see a modification of the neighbour graph

technique[20] to schedule handoff more efficiently in 802.11 type networks, by looking

up the location of the nearest access point, so that channel allocation and preregistration

can be finished before the handoff actually starts, thus reducing handoff delay

significantly.

- 8 -

Chapter 5 – Overview of Current Methods

As mentioned in the introduction, for a best case scenario, handoff techniques

require to be started at an optimum time, as late as possible, but never too late.

Also, the handoff algorithm should, for it's own usage, use up as little of the

resources as is feasible for acceptable performance.

In most currently deployed solutions, one of the above comes at a very

unpleasant lack of the other. Thus, a wasteful tradeoff is executed, resulting in a sub-

optimal handoff performance.

At the current rate of growth of wirelessly connected devices, where we move to

Everything over IP (EoIP), conservative estimated suggest a trillion connected devices

by 2050, compared to around 4-5 billion now. Just we can see that there will be nearly

no scope for bandwidth wastage, and further optimization has to occur.

Here, we shall investigate the performance of some of the current production

and prototype algorithms for handoff, and discuss their shortcomings and advantages.

We shall also analyse the benefit that can be obtained if algorithmic modifications and

novel solutions suggested by academia and industry, if incorporated into current or new

systems can improve their handoff performance considerably. We shall give special

emphasis to methods that use location information of some sort as part of the handoff

management.

5.1 Standard Handoff mechanisms in IEEE 802.11

First, as a point of easy reference to the suggested modifications, we mention the

typical algorithms that are currently under use in common networks, in the following

order : Wi-fi and LTE.

- 9 -

Fig 5.1 The typical IEEE 802.11 Handoff procedure [21]

5.2 Standard Handoff mechanisms in LTE networks

Fig 5.2 shows the basic structure of call handoff [22] when the control plane is

already established, and there will only be a change in the user plane. Thus there will be

no change in the Serving Gateway(SGW). The handoff taking into consideration the

SGW change, called the X2 handover, is a more general case.

- 10 -

Fig 5.2 . Handoff flow in LTE without SGW change [23]

Fig 5.3 LTE X2 Based Handover [23]

Fig 5.3 shows a more complete handoff mechanism, where X2 is the interface

between two eNB’s and X2-AP is the protocols used for communication over it.

Handoff takes place when the eNB detects that UE can no longer be served by it

because of the power constraints.

- 11 -

Chapter 6 – Suggested improvements

As is seen in Chapter 4, handoff mangement has been rigorously investigated. In this

chapter two possible improvements are suggested to the handoff mechanism, assuming

location data is available.

6.1 Modified Neighbour Graph Technique for IEEE 802.11

First we show the neighbour graph method[20] flow diagram in Fig. 6.1.

Fig 6.1 Neighbour Graph flow diagram

Thus we see that for the neighbour graph to be correctly generated, we need to

actively scan for all the adjoining access points. For a homogenous honeycomb

network, this figure is 6 cells/ APs. If location information and cell map is available, we

can reduce the number of scans to be made significantly. Two cases of MS position are

considered, which can be generalised to represent signal state in all cellular systems, at

any positions.

Case A, with radial velocity of MS towards one of the corners of the hexagonal

cell.

Case B, with radial velocity of MS towards one of the flat edges of the

hexagonal cell.

- 12 -

Fig 6.2 Case A with 2 adjoining NBS

Fig 6.3 Case B with 3 adjoining NBS

6.2 Pre-Selective Scanning with GPS in IEEE 802.11

As we can see in Fig. 5.1, handoff delay in 802.11 networks consists of probe

delay,authentication delay and reassociation delay.

We are proposing a scheme named “Pre-Selective Scanning with GPS” to

reduce the scanning(probe) delay which is 90 percent of the whole handoff delay.

Our scheme has the advantage of selective scanning. There are 14 channels used

to be scanned. But only first 11 channels of them are used by IEEE 802.11b. But 1,6,11

- 13 -

number channels are not overlapping; therefore most of APs are likely to operate on

these three channels. We also use pre-scanning mechanism where we can complete the

scanning phase almost before handoff such that scanning delay will not be included in

handoff delay.

In this way, we can reduce handoff delay drastically. We also use GPS to

calculate the distance between the Mobile Node (MN) and the best AP in the old AP’s

neighbourhood area when the MN moves within old base station’s cell. We describe our

proposed work in flowchart form in Fig 6.4

Taking,

n = no of times cache is updated with neighbour graph, N=no.of potential channels.

Let us assume, actual call time before handoff needed is t. So N=t

T s. If our

scheme is applied, then one time will come when n will be just less than 1, in that time

the value of n will be:

n=N−0.005×[(200×(N−1))]

And the scanning delay will be

Sd=T s−t+.005×T s×[(200×(tT s

−1))]

To simulate our proposed algorithm, we have found a formula to find scanning

delay as function of actual call time before handover. So at every interval of time

T s

100, minimum change of n may be 0, and maximum change of n is 0.1.

So, Avg. change of n=(0+.01+.02+.03+.04+.05+.06+.07+.08+.09+0.1)

11=.05

So at every interval of time T s

100, n can be changed by 0.05 on average.

For choosing the value of T s we can have different combination of values.

Pre scanning scheme Value of T s (in ms) Reference

Basic active scan 332.00 [24]

Fast active scan 32.36 [25]

Basic selective scan 129.00 [26]

Fast selective scan 3.00 [27]

- 14 -

Fig 6.4 Pre-Selective Scanning Algorithm

- 15 -

Chapter 7 - Results & Discussion

7.1 Modified Neighbour Graph Technique

It has been found experimentally that each adjoining NBS takes atleast 3ms to

probe. This takes the total probe time to (3×6)=18ms . However by accessing the

cross layer mobility information we can get an estimate of the RSS[11] and velocity[10] of

propagation of the mobile station.

According to Narsimhan & Cox[10] [11], we can only consider radial velocity for

Received Signal Strength velocity estimation using Doppler effect. If the mobile station

moves along one of the edges of the hexagonal honeycomb cell [Case A], then the

probability of moving to an adjoining new base station gets reduced from six to two, as

with radial velocity in a low latency wireless network movement into any other new

base station (NBS) other than these two is geometrically impossible. This drastically

reduces the probe delay minimum channel time from

(6×3)=18ms to (2×3)=6ms.

Similarly, for Case B, If the mobile station does not move along the edge of a

hexagonal honeycomb radially then for a low latency wireless area network it can move

into only 3 adjoining cells by using the same logic as above. In this scenario the

minimum probe time reduces to (3×3)=9ms .

Thus we see that in both cases the probe delay time is reduced by

(18−9)=9ms and in the other situation it is reduced by (18−6)=12 ms by using

selective scanning technique in Neighbour Graph method. Hence probe delay, which

constitutes typically 90% of the total delay can be reduced by upto 67%, thus greatly

improving handoff performance.

- 16 -

7.2 Pre-Selective Scanning with GPS

So, from Fig 7.1 to 7.4, we can realize that the effective delays are lesser than

the previous value because of pre-scanning. The simulation results indicate that the

average delays for basic active scanning, fast active scanning, basic selective scanning

and selective scanning using cache are largely minimised. So if we use our proposed

pre-scanning mechanism, handoff delay can be reduced to a great extent.

Fig 7.1 Scanning delay (basic active scan) Fig 7.2 Scanning delay (fast active scan)( T s = 332 ms) ( T s = 32.36 ms)

Fig 7.3 Scanning delay (basic selective Fig 7.4 Scanning delay(selective scanning)scanning) ( T s = 141 ms) using cache ( T s = 3 ms)

Pre scanning schemeT s (in ms) with normal scan

Effective value of delay after pre-scanning (in ms)

Basic active scan 332.00 0.825

Fast active scan 32.36 0.08

Basic selective scan 129.00 0.35

Fast selective scan 3.00 0.0075

So, we see a marked improvement in handoff delay minimisation.

- 17 -

Chapter 8 – Conclusion and Future Scope

We can see that because of it's inherent role in the way cellular communication

systems function, handoff management is a primary area of study in the design of all

such telecommunication networks and systems.

The utilisation of location information has been broadly discussed and available

literature, some of which is discussed here, makes us expect better performance in terms

of minimisation of handoff delay, hand off failure and false handoff initiation.

This, however, comes at the price of complexity, poorer battery performance

(GPS and velocity estimation techniques are power guzzlers), and lack of reliability.

The last point arises from the issue of GPS being under an unilateral military

administration, thus causing reasonable lack of independent calibration, fine tuning and

technical support. Also, it works only in an outdoor low-cloud environment, with errors

that can go upto 50 m, thus rendering them unsuitable for heterogeneous and

microcellular networks. Also the current unavailability of GPS receivers or similar

equipment universally results in our being unable to deploy it within the standard. The

aforementioned environmental limitations also suggest that location information may

not be available at all times even in mobile stations equipped with relevant hardware

and software.

So we can say that while the availability of location information improves the

efficiency of the handoff mechanism, a lot of work remains to be done to ensure its

universal availability. This leaves scope for a lot of future research and study, for

development of reliable, accessible and uniform acquisition of such relevant

information.

- 18 -

REFERENCES

(1) “Cellular Communication”, Britannica Academic Edition.

(2) “Handoff”, Wikipedia, the free encyclopedia (retrieved Apr 6, 2011).

(3) Viterbi, A.J.; Viterbi, A.M.; Gilhousen, K.S.; Zehavi, E.; , "Soft handoff

extends CDMA cell coverage and increases reverse link capacity," Selected

Areas in Communications, IEEE Journal on , vol.12, no.8, pp.1281-1288, Oct

1994.

(4) Israt, P.; Chakma, N.; Hashem, M.; , "A fuzzy logic-based adaptive handoff

management protocol for next-generation wireless systems," Computer and

Information Technology, 2008. ICCIT 2008. 11th International Conference on ,

vol., no., pp.288-293, 24-27 Dec. 2008

(5) Ian F. Akyidiz, Shantidev Mohanty and Jiang Xie, “A Ubiquitous Mobile

Communication Architecture for Next-Generation Heterogeneous Wireless

Systems”, IEEE Radio Communications, June 2005.

(6) “IEEE Standard for Information technology — Telecommunications and

information exchange between systems — Local and metropolitan area

networks — Specific requirements — Part 11: Wireless LAN Medium Access

Control (MAC) and Physical Layer (PHY) Specifications”, IEEE 802.11-2007.

(7) Shantidev Mohanty, Ian F. Akyildiz, "A Cross-Layer (Layer 2 + 3) Handoff

Management Protocol for Next-Generation Wireless Systems," IEEE

Transactions on Mobile Computing, vol. 5, no. 10, pp. 1347-1360, Oct. 2006.

(8) Shantidev Mohanty, “Architecture and Cross-Layer Mobility Management

Protocols for Next-Generation Wireless Systems”, Ph.D. Thesis, Georgia.

Tech., 2005.

(9) Jamalipour, A.; Tekinay, S.; , "Guest editorial - fourth generation wireless

networks and interconnecting standards," Personal Communications, IEEE ,

vol.8, no.5, pp.8-9, October 2001.

- 19 -

(10) Ravi Narasimhan and Donald C. Cox, ”Speed Estimation in Wireless

Systems Using Wavelets”, Cox, IEEE Transactions on Communications, Vol.

47, No. 9, September 1999.

(11) Ravi Narasimhan and Donald C. Cox, “Estimation of Mobile Speed and

Average Received Power in Wireless Systems Using Best Basis Methods”Cox,

IEEE Transactions on Communications, VOL. 49, NO. 12, December 2001.

(12) Koushik Kumar Nundy, Himanish Mukherjee, Debabrata Sarddar, M. K.

Naskar, U.Biswas, “Minimisation of Hand-off Failure Probability and False

Hand-off Initiation in Low Latency Next Generation Wireless Networks”, Int.

Jnl. of Information Technology and Knowledge Management, Vol 3 No 1, Jan-

Jul 2010.

(13) Debabrata Sarddar, Joydeep Banerjee, Jaydeep Ghosh Chowdhury, Ramesh

Jana, Koushik Kumar Nundy, Utpal Biswas, M.K.Naskar, “Reducing Handoff

delay by Pre-selective Scanning using GPS” ,Int.. Jnl. of Distributed and

Parallel Systems Vol.1, No.2, Nov 2010, pp.142-151.

(14) Huang et al, “Handoff monitoring in cellular communication networks

using slow frequency hopping”, United States Patent. 5,448,569 Sep. 5, 1995.

(15) Dutta, Ashutosh; Madhani, Sunil; Wai Chen; Altintas, Onur; Schulzrinne,

Henning; , "GPS Assisted Fast-Handoff Mechanism for Real-Time

Communication," Sarnoff Symposium, 2006 IEEE , vol., no., pp.1-4, 27-28

March 2006.

(16) Celebi, H.; Arslan, H.; , "Utilization of Location Information in Cognitive

Wireless Networks," Wireless Communications, IEEE , vol.14, no.4, pp.6-13,

August 2007.

(17) Cortes-Rodriguez, F.; Munoz-Rodriguez, D.; Soto, R.; , "Position location

assisted multi-valued logic handoff algorithm," Vehicular Technology

Conference, 1999. VTC 1999 - Fall. IEEE VTS 50th , vol.2, no., pp.775-779

vol.2, 1999.

(18) AK Raith, "Position assisted handoff within a wireless communications

network", US Patent 6,711,408, 2004.

- 20 -

(19) Koushik Kumar Nundy, Himanish Mukherjee, Debabrata Sarddar, M. K.

Naskar, U.Biswas, “Minimisation of Handoff delay in 802.11 WLAN networks

using modified neighbour graph method.”, under review.

(20) S. Pack, H. Jung, T. Kwon and Y. Choi, “A selective neighbor caching

scheme for fast handoff in IEEE 802.11 wireless networks”, IEEE

Communications, IEEE International Conference, Volume 5, Page(s):3599 –

3603, Vol. 5, May 2005.

(21) Arunesh Mishra, Minho Shin, and William Arbaugh. 2003. “An empirical

analysis of the IEEE 802.11 MAC layer handoff process.” SIGCOMM Comput.

Commun. Rev. 33, 2 (April 2003).

(22) Kyungmin Kim, “Voice Call Handover Scheme between 3G LTE and 3G

CS Network”, M.Sc Thesis, Yonsei University, Dec 2007.

(23) S. K. Dornal, “LTE Whitepaper”, 3G4G Wireless Resource Centre, UK,

2009.

(24) S. Shin, A.S. Rawat, H. Schulzrinne. “Reducing MAC layer Handoff

Latency in IEEE 802.11 Wireless LANs”. In proceedings of 2nd International

Workshop on Mobility Management and Wireless Access Protocol, 2004, pp

19-26.

(25) J. Pesola and S. Pokanen, “Location-aided Handover in Heterogeneous

Wireless Networks,” Proceedings of Mobile Location Workshop, May 2003.

(26) H.S. Kim, S.H. Park, C.S. Park, J.W. Kim, S.J. Ko. “Selective Channel

Scanning for Fast Handoff in Wireless LAN using Neighbor Graph”. In

proceedings of the 9th International Conference on Personal Wireless

Communication, 2004, pp.=194-203.

(27) Dutta, A.; Chakravarty, S.; Taniuchi, K.; Fajardo, V.; Ohba, Y.; Famolari,

D.; Schulzrinne, H.; , "An Experimental Study of Location Assisted Proactive

Handover,"Global Telecommunications Conference, 2007. GLOBECOM '07.

IEEE, vol., no., pp.2037-2042, 26-30 Nov. 2007.

(28) Lau, S.S.-F.; Kwan-Fai Cheung; Chuang, J.C.I.; , "Fuzzy logic adaptive

handoff algorithm," Global Telecommunications Conference, 1995.

GLOBECOM '95., IEEE , vol.1, no., pp.509-513 vol.1, 14-16 Nov 1995.

- 21 -

(29) Syuhada, M.Z.A.; Mahamod, I.; Firuz, W.A.W.N.S.; , "Performance

Evaluation of Vertical Handoff in Fourth Generation (4G) Networks Model,"

Telecommunication Technologies 2008 and 2008 2nd Malaysia Conference on

Photonics. NCTT-MCP 2008. 6th National Conference on , vol., no., pp.392-

398, 26-28 Aug. 2008.

(30) Chien-Chao Tseng; Li-Hsing Yen; Hung-Hsin Chang; Kai-Cheng Hsu; ,

"Topology-aided cross-layer fast handoff designs for IEEE 802.11/mobile IP

environments," Communications Magazine, IEEE , vol.43, no.12, pp. 156- 163,

Dec. 2005.